Summary and Abstract- The main goals of this proposal are 1) to provide the fundamental knowledge required for understanding Cytochrome P450 mediated reaction mechanism with regards to rates, regioselectivity, and binding, and 2) to provide computational tools for predicting the metabolic component of ADMET (Absorption, Distribution, Metabolism, Excretion, Toxicity). These are highly significant goals that will positively impact almost all drug development projects and decrease the time required to develop new therapeutics. These goals will be described in terms of the following four Specific Aims: Aim 1) At present the number of active-oxygen species used by Cytochrome P450 remains controversial, and we propose that this may explain why predictive methods for metabolism are limited to around 85% accuracy. Predictive models for metabolism are important in drug design, and have the potential to save lives by decreasing the time it takes to develop new drugs. Aim 2) Specific Aim 2 explores binding afforded to substrates designed to coordinate to the iron of the heme (type II binding) providing specificity for P450 enzymes by increasing affinity up to 250-fold. This is significant because systemic administration of drugs meant to inhibit a single P450 enzyme normally leads to broad inhibition of a number of P450 enzymes, upsetting homeostasis, and causing drug-drug interactions. Aim 3) Computational prediction of P450 mediated rates remains one of the most important targets of researchers working in the field. Specific Aim 3 will establish the features important in such predictions. If this goal can be met we can understand the important features involved in predicting the clearance of a drug from the body, and we move closer to the goal of virtual drug design. Aim 4) With the ever-expanding number of computational methods and research publications in predictive ADMET, Specific Aim 4 will provide common sets of publicly accessible test set databases along with open-source predictive code allowing different computational ADMET methods to be tested against common benchmarks. This is particularly important since the majority of the methods are developed in-house and the test sets are not published. We hypothesize that by publishing open source code to a public web site for metabolic predictions that we can translate the results of ours' and others' research into the public domain resulting in a significant increase in the use and quality of these tools. We also hypothesize that having a common set of data available to all researchers will encourage validation, comparison, and enhancement of ADMET models. [unreadable] [unreadable] Public Health Relevance: Narrative- The purpose of this grant application is to understand the important features of cytochrome P450 enzymes with respect to drug metabolism, and drug design. Cytochrome P450 enzymes are the most important drug metabolizing enzymes and are responsible for most drug metabolism. While the majority of the reactions mediated by this enzymes family are benign, a number cause activation to reactive species that can cause cancer and toxicity. Furthermore, many life threatening drug-drug interactions occur from drugs slowing cytochrome P450 mediated reactions. This grant application will develop methods to design new drugs faster, and safer than we can presently through an increased understanding of the rates and binding affinities or P450 mediated reactions. [unreadable] [unreadable] [unreadable]